Policy makers, program managers, community planning group members and other key decision makers need to balance the costs and effectiveness of various interventions when planning and evaluating HIV prevention programs. Resources to fund these programs are limited, and must be used judiciously in order to maximize the number of HIV infections averted. Economic evaluation studies of HIV prevention interventions can provide source of the information needed by key decision makers. Ideally, decision makers would have access to information on the cost- effectiveness of multiple types of prevention interventions as delivered to several populations. Given the convincing data that cognitive- behavioral interventions small-group and community level interventions can modify or avert HIV-related risk behaviors, decision makers must consider whether and how to mix such interventions int their current program portfolios. When making resource allocation decisions about behavioral interventions, decision makers need to know the costs and cost-effectiveness of the programs. Unfortunately, little such information is currently available. A recent review found only 28 economic evaluation studies of HIV prevention interventions (in the U.S.) which included any consideration at all of behavior change issues. Of these studies, 16 focused on counseling and testing services (for several populations), and 13 focused on several types of interventions for injection drug users. Cognitive-behavioral small-group and community- level interventions have received essentially no attention. In this proposal, a research effort is described to perform economic evaluation studies of cognitive-behavioral small-group and community-level interventions for three important populations: women in high risk situations, gay men and adolescents. Because effectiveness data have been published for these types of interventions, these economic evaluation studies will be retrospective. The cost of the interventions will be estimated retrospectively using published information about intervention delivery. The published behavioral outcome data will be mapped into health outcomes (in particular, estimated number of HIV infections averted) using mathematical modelling techniques. The estimated cost and health outcomes for each intervention, for each specific population, will be combined using standard methods of cost-utility analysis (CUA). CUAs contain estimates of how much the intervention cost for each quality adjusted life year (QALY) saved. The cost-per-QALY outcome measure is a commonly accepted one, includes consideration of morbidity and mortality, and allows comparisons of HIV prevention interventions to health service programs from other, non-HIV-related areas. The proposed research program will include publication of the study results in scientific journals, but will also include a special dissemination plan for both methods and results. Each year a guide to selected methodological topics in economic evaluation will be published for an audience of behavioral and social scientists doing HIV prevention research (in the hope of making their research even more policy-relevant). Also, a special package of the published articles and guides will be assembled, and an interpretive overview written for multiple audiences of decision makers.